[DBSCAN scikit python eps意外

问题描述 投票:-1回答:1

我已经检查了DBSCAN scikit问题(已经很老了,但是我的代码是给出错误:

[DBSCAN()得到一个意外的参数eps

输入的不是我的实际输入,只是测试值,但是我现在有这个问题。如果您能提供帮助,我将不胜感激

from sklearn.cluster import DBSCAN
import numpy as np

def clusterCenter(ll:list):
    x=0
    y=0
    for el in ll:
        x=x+el[0]
        y=y+el[1]
    x=x/len(ll)
    y=y/len(ll)    
    return (x,y)

def printCoords(clist:list):
    for el in clist:
        x,y = el
        print( str(x)+", "+str(y) )

def DBSCAN(X:np.array, max_distance:float, min_nodes_in_cluster:int):
    clusterCenters=[]
    dbscan = DBSCAN(eps=max_distance, min_samples=min_nodes_in_cluster,metric="euclidian").fit(X)
    alllabels=dbscan.labels_

    #getting the points in each cluster by using a dictionary
    num=0
    dict={}
    for el in alllabels:
        if dict.get(el) is None:
            somearr=[]
            somearr.append(X[num])
            dict[el]=somearr
        else:
            dict[el].append(X[num])
        num=num+1

    #print all cluster centers for each cluster
    for key in dict:
        x,y=clusterCenter(dict[key])
        clusterCenters.append((x,y))
    return clusterCenters

X = np.array([[1, 2], [2, 2], [2, 3],[8, 7], [8, 8], [25, 80]])
cclist=DBSCAN(X,1.3,5)
printCoords(cclist)
python dbscan
1个回答
0
投票

问题是您将函数定义为DBSCAN,它将覆盖从sklearn.cluster导入的名称。只需将函数重命名为其他名称即可。

© www.soinside.com 2019 - 2024. All rights reserved.